Expression Patterns of Coagulation Factor XIII Subunit A on Leukemic Lymphoblasts Correlate with Clinical Outcome and Genetic Subtypes in Childhood B-cell Progenitor Acute Lymphoblastic Leukemia

Abstract

Background: Based on previous retrospective results, we investigated the association of coagulation FXIII subunit A (FXIII-A) expression pattern on survival and correlations with known prognostic factors of B-cell progenitor (BCP) childhood acute lymphoblastic leukemia (ALL) as a pilot study of the prospective multi-center BFM ALL-IC 2009 clinical trial.

Methods: The study included four national centers (n = 408). Immunophenotyping by flow cytometry and cytogenetic analysis were performed by standard methods. Copy number alteration was studied in a subset of patients (n = 59). Survival rates were estimated by Kaplan-Meier analysis. Correlations between FXIII-A expression patterns and risk factors were investigated with Cox and logistic regression models.

Results: Three different patterns of FXIII-A expression were observed: negative (<20%), dim (20-79%), and bright (≥80%). The FXIII-A dim expression group had significantly higher 5-year event-free survival (EFS) (93%) than the FXIII-A negative (70%) and FXIII-A bright (61%) groups. Distribution of intermediate genetic risk categories and the "B-other" genetic subgroup differed significantly between the FXIII-A positive and negative groups. Multivariate logistic regression confirmed independent association between the FXIII-A negative expression characteristics and the prevalence of intermediate genetic risk group.

Conclusions: FXIII-A negativity is associated with dismal survival in children with BCP-ALL and is an indicator for the presence of unfavorable genetic alterations.

Keywords: B-cell progenitor; Factor XIII subunit A; acute lymphoblastic leukemia; children; genetic risk categories; middle-income countries; survival.

Figure 1

Representative dot plots and histograms of leukemic lymphoblasts. There are three different patterns of cytoplasmic FXIII-A expression in terms of positivity of leukemic lymphoblasts (CD45 negative population): (I) negative expression pattern below 20%, (II) dim positive expression pattern between 20% and 79%, and (III) bright positive expression pattern ≥ 80% of leukemic lymphoblasts with an FXIII-A staining exceeding the intensity of negative controls, i.e., residual normal lymphocytes (grey). Based on FXIII-A expression intensity of normal residual lymphocytes, the threshold of positivity is marked by the dashed line on the respective dot-plots. Percentages are referring to lymphoblasts. The intensity of the FXIII-A expression increased continuously, as the histogram of lymphoblasts with dim expression pattern shows (indicated by blue colour), which excludes the existence of a distinct FXIII-A negative (green colour) and a FXIII-A bright (red colour) subpopulation.

Figure 2

Ratio of FXIII-A expression of leukemic lymphoblast on Day 0 and Day 15. Thirty-six patients had FXIII-A expression data on Day 15 (a). Leukemic lymphoblasts were FXIII-A positive in 27 cases (b) and FXIII-A negative by 9 patients (c). FXIII-A expression of negative blasts was less than 20% (dashed line). p < 0.05 values were considered statistically significant, such p values are marked in bold.

Figure 3

Event-free and overall survival of children with B-cell precursor acute lymphoblastic leukemia (ALL) by FXIII-A expression pattern. According to the Kaplan-Meier analysis, patients with FXIII-A dim lymphoblasts (blue line) had significantly higher 5-year event-free survival (EFS) (93%) compared with patients with FXIII-A bright lymphoblasts (red line; 61%) and with FXIII-A negative lymphoblasts (green line; 70%) (a). Mean follow-up time for the FXIII dim group was 1736 days (95% CI: 1675 and 1796 days); for the FXIII-A bright group 1277 days (95% CI: 1153 and 1400 days); and for the FXIII-A negative group 1588 days (95% CI: 1484 and 1692 days), with the last censored event at 31 days. Difference of 5-year OS was significant between patients with FXIII-A dim lymphoblasts (blue line; 95%) and patients with FXIII-A negative lymphoblasts (green line; 88%). Difference between the 5-year OS of the FXIII-A dim group and the FXIII-A bright group (red line; 87%) was not significant (b). Mean follow-up time for the FXIII-A dim group was 1755 days (95% CI: 1699 and 1810 days); for the FXIII-A bright group 1454 days (95% CI: 1341 and 1567 days); and for the FXIII-A negative group 1661 days (95% CI: 1572 and 1749 days), with the last censored event at 31 days. Statistically significant differences between the respective groups are marked with arrows (a,b). Shaded regions indicate that at any time point, there is a 95% chance that the interval contains the true percentage survival.

Figure 4

Detection of FXIII-A expression on analysis of serial dilutions. Based on a three parallel serial dilutions (10×; 100×; 1000×), the percentage of FXIII-A expression in residual lymphoblasts could be clearly assessed when the percentage of lymphoblasts was above 0.04%. Black color indicates the residual lymphoblasts, grey color represents the mature lymphocytes, which were the negative control. (a) The percentages of FXIII-A positive lymphoblasts within all lymphoblasts were similar in the three parallel staining instances (b) The percentages of lymphoblasts served as the internal control of the serial dilutions (c).